Resource Efficient Gadgets for Compiling Adiabatic Quantum Optimization Problems

Abstract

A resource efficient method by which the ground-state of an arbitrary k-local, optimization Hamiltonian can be encoded as the ground-state of a inline image-local, optimization Hamiltonian is developed. This result is important because adiabatic quantum algorithms are often most easily formulated using many-body interactions but experimentally available interactions are generally 2-body. In this context, the efficiency of a reduction gadget is measured by the number of ancilla qubits required as well as the amount of control precision needed to implement the resulting Hamiltonian. First, methods of applying these gadgets to obtain 2-local Hamiltonians using the least possible number of ancilla qubits are optimized. Next, a novel reduction gadget which minimizes control precision and a heuristic which uses this gadget to compile 3-local problems with a significant reduction in control precision are shown. Finally, numerics are presented which indicate a substantial decrease in the resources required to implement randomly generated, 3-body optimization Hamiltonians when compared to other methods in the literature.